Evaluating empiric antibiotic prescribing for hospitalized children in Mozambique through the introduction of a quarterly syndromic antibiogram: An implementation science protocol

Antimicrobials are the most frequently prescribed drug in pediatrics, with an estimated 37% of infants and 61% of hospitalized children having received them. Approximately 20-50% of prescriptions have been shown to be potentially unnecessary or inappropriate. The World Health Organization (WHO) estimates that the continued increase in antimicrobial resistance by the year 2050 will lead to the death of 10 million people per year. This paper describes a protocol to be used in a future study to evaluate the implementation of a quarterly syndromic antibiogram, aimed to improve the use of antibiotics for the treatment of pediatric bacterial infections at the Maputo Central Hospital, Mozambique. This study uses implementation science methods framed by the Dynamic Adaption Process (DAP) and RE-AIM conceptual frameworks to develop a multi-phase, mixed-methods evaluation utilizing qualitative and quantitative approaches. The pediatric inpatient services at HCM consist of approximately 18 physicians and 60 nurses. Additionally, the microbiology laboratory consists of eight laboratory technicians. We anticipate analyzing approximately 9,000 medical records. Qualitative methods include in-depth interviews with clinicians, laboratory technicians, and administrators to explore current knowledge and practices around antibiotic decision making, facilitators and barriers to intervention implementation, as well as acceptability and satisfaction with the intervention roll-out. Qualitative analysis will be performed with NVivo 12 software. Quantitative methods include extracting data from existing records from the pediatric ward of Hospital Central de Maputo (HCM) guided by the RE-AIM framework to explore intervention utilization and other factors influencing its implementation. Quantitative descriptive and inferential statistical analysis will be performed using R Studio statistical software. The findings from this evaluation will be shared with hospital administrators and relevant national policymakers and may be used by the Ministry of Health in deciding to expand this approach to other hospitals. The expected results of this research include the development of standard operating guidelines for the creation, distribution, and use of a quarterly syndromic antibiogram for antibiotic decision making that is informed by local epidemiology. Findings from this study will be used to develop a larger multi-site trial in Mozambique.